A deliberative framework to identify the need for real-life evidence building of new cancer drugs after interim funding decision.

BACKGROUND: With the rising cost of new oncology treatments, it is no longer sustainable to base initial drug funding decisions primarily on prospective clinical trials as their performance in real-life populations are often difficult to determine. In British Columbia, an approach in evidence building is to retrospectively analyse patient outcomes using observational research on an ad hoc basis. METHODS: The deliberative framework was constructed in three stages: framework design, framework validation and treatment programme characterization, and key informant interview. Framework design was informed through a literature review and analyses of provincial and national decision-making processes. Treatment programmes funded between 2010 and 2013 were used for framework validation. A selection concordance rate of 80% amongst three reviewers was considered to be a validation of the framework. Key informant interviews were conducted to determine the utility of this deliberative framework. RESULTS: A multi-domain deliberative framework with 15 assessment parameters was developed. A selection concordance rate of 84.2% was achieved for content validation of the framework. Nine treatment programmes from five different tumour groups were selected for retrospective outcomes analysis. Five contributory factors to funding uncertainties were identified. Key informants agreed that the framework is a comprehensive tool that targets the key areas involved in the funding decision-making process. CONCLUSIONS: The oncology-based deliberative framework can be routinely used to assess treatment programmes from the major tumour sites for retrospective outcomes analysis. Key informants indicate this is a value-added tool and will provide insight to the current prospective funding model.

COST-EFFECTIVENESS IMPACTS CANCER CARE FUNDING DECISIONS IN BRITISH COLUMBIA, CANADA, EVIDENCE FROM 1998 TO 2008.

OBJECTIVES: The Priorities and Evaluation Committee (PEC) funding recommendations for new cancer drugs in British Columbia, Canada have been based on both clinical and economic evidence. The British Columbia Ministry of Health makes funding decisions. We assessed the association between cost-effectiveness of cancer drugs considered from 1998 to 2008 and the subsequent funding decisions. METHODS: All proposals submitted to the PEC between 1998 and 2008 were reviewed, and the association between cost-effectiveness and funding decisions was examined by (i) using logistic regression to test the hypothesis that interventions with higher incremental cost-effectiveness ratios (ICERs) have a lower probability of receiving a positive funding decision and (ii) using parametric and nonparametric tests to determine if a statistically significant difference exists between the mean cost-effectiveness of funded versus not funded proposals. A sub-analysis was conducted to determine if the findings varied across different outcome measures. RESULTS: Of the 149 proposals reviewed, 78 reported cost-effectiveness using various outcome measures. In the proposals that used life-years gained as the outcome (n = 22), a statistically significant difference of nearly $115,000 was observed between the mean ICERs for funded proposals ($42,006) and for unfunded proposals ($156,967). An odds ratio indicating higher ICERs have a lower probability of being funded was also found to be statistically significant (p < .05). CONCLUSIONS: Economic evidence appears to play a role in British Columbia cancer funding decisions from 1998 to 2008; other decision-making criteria may also have an important role in recommendations and subsequent funding decisions.

Adherence with capecitabine: A population-based analysis based on prescription refill data.

Background Patient adherence is important with the increasing use of oral anticancer drugs. Recent studies reported different capecitabine adherence rates based on self-reporting and microelectronic monitoring of the medication bottle. Patient's awareness of being monitored may confound these results. Prescription records provide a larger and more objective dataset for adherence investigation. We report the use of computer algorithm and manual review of prescription and medical documentation to determine the rate of capecitabine adherence. Methods Two years of capecitabine prescription records from five ambulatory cancer centres were reviewed. Prescription data were extracted using a custom Java-based software tool to compare the predicted vs. actual dispensing date. The difference between the dates was the primary adherence measure (altered treatment date incident) and estimated using a computer algorithm and by manual review of medical charts. Results Of 4412 refill prescriptions, 45.2% was associated with an altered treatment date incident based on the initial computer algorithm. This was reduced to 29.5% after adjusting for clinic scheduling processes and 10.2% after manual chart review to adjust for valid reasons for delay. The reasons for altered treatment date incident were not identified in 52.2% of prescriptions. Conclusions Adherence rate of capecitabine based on refill data seem to be high and consistent with other findings based on patient self-report. Population analysis of prescription data with custom computer algorithm may identify trends in capecitabine adherence with some efficiency. Manual review would likely be required to verify these results. The accuracy of using altered prescription refill dates as an adherence measure requires further studies.

A Time-Trend Economic Analysis of Cancer Drug Trials.

BACKGROUND: Scientific advances have led to the discovery of novel treatments with high prices. The cost to publicly fund high-cost drugs may threaten the sustainability of drug budgets in different health care systems. In oncology, there are concerns that health-benefit gains are diminishing over time and that the economic evidence to support funding decisions is too limited. METHODS: To assess the additional costs and benefits gained from oncology drugs over time, we used treatment protocols and efficacy results from U.S. Food and Drug Administration records to calculate cost-effectiveness ratios for drugs approved to treat first- and second-line metastatic or advanced breast, colorectal, and non-small cell lung cancer during the years 1994-2013. We assessed reimbursement recommendations reached by health technology assessment agencies in the U.K., Australia, and Canada. RESULTS: Cost-effectiveness ratios were calculated for 50 drugs approved by the U.S. regulator. The more recent approvals were often based on surrogate efficacy outcomes and had extremely high costs, often triple the costs of drugs approved in previous years. Over time, the effectiveness gains have increased for some cancer indications; however, for other indications (non-small cell lung and second-line colorectal cancer), the magnitude of gains in effectiveness decreased. Reimbursement recommendations for drugs with the highest cost-effectiveness ratios were the most inconsistent. CONCLUSION: Evaluation of the clinical benefits that oncology drugs offer as a function of their cost has become highly complex, and for some clinical indications, health benefits are diminishing over time. There is an urgent need for better economic evidence from oncology drug trials and systematic processes to inform funding decisions. IMPLICATIONS FOR PRACTICE: High-cost oncology drugs may threaten the ability of health care systems to provide access to promising new drugs for patients. In order to make better drug-funding decisions and enable equitable access to breakthrough treatments, discussions in the oncology community should include economic evidence. This study summarizes the extra benefits and costs of newly approved drugs from pivotal trials during the postgenomic era of drug discovery. The reader will gain an appreciation of the need for economic evidence to make better drug-reimbursement decisions and the dynamics at play in today's oncology drug market.

Incremental cost-effectiveness of the pre- and post-bevacizumab eras of metastatic colorectal cancer therapy in British Columbia, Canada.

OBJECTIVES: Bevacizumab, with 5-Fluorouracil-based therapy, has been shown to prolong survival among patients with metastatic colorectal cancer (mCRC), at a relatively high cost. We evaluated the cost-effectiveness of therapy for mCRC in the eras prior to, and following, the introduction of bevacizumab on a population basis. METHODS: All patients with newly-diagnosed mCRC in 2003/2004 (the pre-bevacizumab era) and 2006 (the post-bevacizumab era) referred to the British Columbia Cancer Agency were included. A cost-utility/cost-effectiveness analysis was conducted using 'real-world' data and a Markov approach. Costs were estimated from provincial pharmacy and radiation databases, as well as national inpatient, day-surgery and ambulatory care data sources. Cost per quality adjusted life year gained (QALY) and life year gained (LYG) ratios were calculated, and one-way and probabilistic sensitivity analyses were conducted to test the robustness of the model. RESULTS: Median overall survival improved from 15.6 to 19.5 months (p = 0.03). For patients in the post-bevacizumab era there was a gain of an average of 0.06 QALYs or 0.325 LYG at a cost increase of $3791 per patient, resulting in a incremental cost-effectiveness ratio of $62,469/QALY or $15,617/LYG. This improved to $43,058/QALY or $10,764/LYG when the analysis was restricted to patients potentially eligible for bevacizumab. Results were generally robust to changes in model parameters, but were sensitive to the costs of chemotherapy and to differential utilities used in the Markov model. CONCLUSIONS: The introduction of bevacizumab was associated with increased costs, but cost-effectiveness estimates were in-line with those reported for other cancer therapies.

Assessing the real-world cost-effectiveness of adjuvant trastuzumab in HER-2/neu positive breast cancer.

BACKGROUND: Among women with surgically removed, high-risk HER-2/neu-positive breast cancer, trastuzumab has demonstrated significant improvements in disease-free and overall survival. The objective of this study is to evaluate the cost-effectiveness of the currently recommended 12-month adjuvant protocol of trastuzumab using a Markov modeling approach and real-world cost data. METHODS: A 10-health-state Markov model tracked patients' quarterly transitions between health states in the local and advanced states of breast cancer. Clinical data were obtained from the joint analysis of the National Surgical Adjuvant Breast and Bowel Project and North Central Cancer Treatment Group, as well as from the metastatic study conducted by Norum et al. Clinical outcomes were adjusted for quality of life using utility estimates published in a systematic review. Real cost data were obtained from the British Columbia Cancer Agency and were evaluated from a payer perspective. Costs and utilities were discounted at 5% per year, respectively, for a 28-year time horizon. RESULTS: In the base case analysis, treatment with a 12-month adjuvant trastuzumab regimen resulted in a gain of 1.38 quality-adjusted life years or 1.17 life years gained at a cost of $18,133 per patient. Thus, the cost per QALY gained for the base case is $13,095. Cost per LYG is $15,492. CONCLUSIONS: Over the long term, treatment of HER-2/neu mutation positive breast cancer with a 12-month protocol of trastuzumab in the adjuvant setting is predicted to be cost-effective in a Canadian context.

Look-alike, sound-alike drugs in oncology.

BACKGROUND: Medication errors with oncology drugs can place cancer patients at risk for adverse events or death. Look-alike, sound-alike (LASA) drug names may increase the risk for errors. Published lists of LASA drug names are generally a result of voluntarily reported medication incidents. This study performed a proactive review of the oncology drug formulary from the Cancer Services of the Alberta Health Services for LASA drug pairs. METHODS: The Levenshtein Distance and Bigram Similarity algorithms, same first and last letters, and Lexi-Comp(R) on-line alerts were used to review the outpatient oncology formulary to identify potential LASA generic drug name pairs. RESULTS: indicate there are more potential LASA generic drug name pairs in the oncology formulary than are published in the literature. The risk detection methods used in this study identified unique and common LASA drug pairs. The Bigram Similarity algorithm identified 186 LASA drug pairs from 3320 possible pairs. The Levenshtein Distance algorithm, same first and last letters, and Lexi-Comp(R) methods identified 42, 75, and 38 LASA drug pairs, respectively. Five generic LASA drug pairs were identified in common by all four of the risk determination methods. DISCUSSION: LASA drug pairs identified by three or four methods were considered to provide the highest risk for errors. A step-wise approach to risk reduction, dependent on the number of detection methods identifying a pair, is presented. CONCLUSION: For specialty areas of practice, a proactive system of reviewing LASA drug name pairs may be warranted for increasing medication safety.

Long-term survival in patients with metastatic melanoma treated with DTIC or temozolomide.

BACKGROUND: Patients with metastatic melanoma typically have a poor outcome; however, a small proportion of patients achieve long-term survival (LTS). It is unclear how often LTS is related to sensitivity to chemotherapy. METHODS: All patients with metastatic melanoma treated with either dacarbazine (DTIC) or temozolomide (TMZ) at the British Columbia Cancer Agency (BCCA) from January 1, 1988 to February 1, 2006 were identified through the BCCA pharmacy electronic database, which was then linked to the surveillance and outcomes unit to identify patients with LTS, defined as survival > or =18 months following chemotherapy. RESULTS: In total, 397 patients were treated with either DTIC (n = 349) or TMZ (n = 48) and 43 patients (10.8%) were identified with LTS. Two additional patients with LTS were added prior to 1988 for a total of 45 patients. The 5-year overall and progression-free survival rates for patients with LTS were 33% and 16%, respectively. In total, 16% had a complete response (CR) to chemotherapy, which was the only factor identified that correlated with survival in the multivariate analysis. However, most patients with LTS had an incomplete response to chemotherapy. CONCLUSIONS: LTS occurs in select patients who achieve a CR to chemotherapy. However, this occurs in only a minority of patients and, in most cases, the longer survival is likely the result of indolent disease biology or host factors.

Use of combined androgen blockade for advanced prostate cancer in British Columbia.

OBJECTIVES: Initial androgen deprivation therapy (ADT) for metastatic prostate cancer with combined androgen blockade (luteinizing-hormone releasing hormone agonist [LHRH agonist] plus antiandrogen) is not recommended in British Columbia (BC). However, this is difficult to monitor since ADT includes concurrent antiandrogen for the first month of LHRH agonist to prevent disease flare. We describe the prevalence of CAB use in BC and its financial impact. METHODS: This was a population-based, retrospective analysis. Patients started on LHRH agonist in January 2005 to December 2006 were identified from the BC Cancer Agency database. CAB was defined as greater than 1 month of antiandrogen concurrently with LHRH agonist. Incremental cost of CAB was based on an average 18 months of therapy from the pivotal CAB study. Incremental cost-effectiveness ratio (ICER) was based on life-year gained (LYG) from the Prostate Cancer Trialists' Collaborative Group meta-analysis. Estimated financial impact for 2007-2008 was based on an annual increase by 5.5% in prevalence of prostate cancer in BC. RESULTS: A total of 2751 patients were identified. CAB was used in 607 patients (22%), associated with an incremental cost of CDN$1768 and ICER of CDN$11,220/LYG per patient. Total incremental cost was CDN$1,073,176 and estimated to be CDN$1,398,644 for January 2007 to December 2008. CONCLUSION: Nearly one-quarter of patients were treated with CAB for metastatic prostate cancer even though it was not recommended in BC. Additional cost of CAB use was considerable, at CDN$1768 per patient. With increased prevalence of prostate cancer, this has important budget implication for funding agencies which do to recommend CAB.